Valve playtests the shit out of their games. A lot of Portal's theming is based on their extensive testing culture and the ways they've developed to manipulate the player through testing. For example, the Weighted Companion Cube came about because they wanted a level where you solve puzzles while carrying an object. But the playtesters kept leaving the damn object behind. So Valve's solution was to tell them it was a "character" they had to care for.
And then players wouldn't leave it behind and tried to drag the cube with them for the entire rest of the game. So they gave the Companion Cube a dramatic death scene at the end of the level.
Maybe this used to be true (I am not making a claim one way or the other) but it seems unlikely they would have that approach and end up with their latest offering - Artifact. I realize not every game is going to be a hit but that game doesn't seem to be liked by just about anybody.
It’s not a game that I, a long time Valve fan would have any interest in, but I’ve been told by friends who are into card games that the actual game was fun, it was the awful pricing that killed it.
I played Artifact, it was fun until it died, but a very tough set of lessons for Valve to swallow:
1) Their reputation really has been damaged over no HL3/Dota 2 money grabs.
2) That they can't just publish anything onto Steam anymore with little to no upfront communication.
3) Not all monetisation models are viable - especially with the recent loot-box fiascos. If they wanted a card economy the game should've been free with random cards, with the option to purchase cards to have your own collection.
The past few months they've definitely started to change over the response to Artifact. Dota Underlords was shipped in only 5 months, so it just shows what they're capable of with a little pressure.
It's a game that is very fun for people who are already invested in card games. The pricing was a big issue, but another one was that the game just isn't that fun for more casual players. It's a game that is very sensitive to misplays, has much more visible RNG than other card games (but I'd argue not much more RNG total) and a lot of the game's play patterns push you into lines where the right play is to just do nothing.
Artifact is a case study in how user testing alone goes wrong.
They did extensively test Artifact but user testing will get you stuck in local maxima. You still need someone with a good wholistic vision to make a great product.
I heard they had a lot of feedback along the lines of "I really like this game, but I'm not sure if other people will" (this from a friend who is friends with one of the artifact devs, so take it with a healthy dose of salt).
I support wholistic as an alternate spelling of holistic. You're actually likely to come to a good idea of what it means even if you've had no exposure to real word before.
Teams changed too. Companies try to develop identities and cultures but it's often just the sum of its people.
Valve also has a unique decentralized structure where teams work independently on games, as has been noted many times before with their handbook (not sure how accurate that is today). So it's possible you're just seeing the results of a completely different team.
Reminds me of what CroTeam did with The Talos Principle. They started with puzzles and wrote a story and a world. They had human players for testing in addition to an automated bot, and also many variants of the puzzles have been tried.
I've always assumed BlackMesa was made by Valve for some reason, that's cool. But yes, as someone else said, I was referring to Hunt Down the Freeman. At least a couple of the Youtubers involved made videos about their experiences and how poorly the project was managed.
The release of that script was really personal for me having grown up on Half Life and following many of the theories about the nature of g-man and the crossover between portal and Half Life with the Borealis ship being the primary setting of the script.
It was released after Marc Laidlaw (writer of Half Life 1 and 2) had resigned due to the constant postponing of Half Life 3 and the erosion of Valve's original values as a company.
It was really sad, but also provided some a much needed conclusion. I could finally say "it's over, half life 3 isn't coming". And even if it does eventually get released it won't be made under the vision of its original spectacular writer.
Valve is all about the benjamins now and has been for a long time. They are 100% business model first, game second. I wouldn't be surprised if they no longer had the talent to make HL3.
Portal 1 and 2 have amazing in-game developer commentary modes. Would highly recommend a replay with the commentary enabled to anyone interested in game design at all.
Can't watch with sound at the moment to confirm it's a good video, but this comes up in youtube
I wish more games had dev commentary like this. It's neat to peek into the minds of the developers, and can give good informal lessons on designing games and levels.
I don't know where the parent got their info, but I learned this through the original Portal's in-game "developer commentary" system. If you turn on the dev commentary in Portal they talk about this fact during that level.
I wish they did this test with people who played the Quake series (Quake / Quake II / Quake 3 / Quake Live). They were arguably one of the most popular non-realism based FPS games of their time (they spanned ~25 years).
The big difference between Portal is, Quake is a multiplayer game where you're playing against a human in real time and there's no backup plan of being able to Google an answer if you get stuck. You either practice and get better or get utterly destroyed. It's really a good case for overcoming challenges and persistence.
There's 1v1 matches where you need to memorize maps to such a degree that you can almost navigate them blind folded. You're also dealing with timing multiple items down to the second, predicting where a human opponent will go, mastering each weapon in every scenario to maximize damage output while minimizing damage input and also generally having a high level overview of how the match is going to figure out when it's worth taking risks or play it safe.
That's just the mental side of it too. Then there's needing very good reaction times and dexterity to aim quickly and precisely and each weapon has its own style of aiming (flick shots, precise tracking, projectile prediction, instant hit scan weapons, etc.).
And then there's other variants of the game like 4v4 TDM or capture the flag that share similar challenges as above but now it's amplified because you're playing with a lot of people and you need to make very interesting decisions with powerups.
All of this happens within seconds and becomes second nature once you've played long enough. It's actually almost unbelievable that a human brain can react so quickly and become so well adjusted to navigating a high speed virtual world under pressure. A lot of these Quake games were played at live events with hundreds of thousands of dollars on the line before e-sports became as huge as it is today.
I still maintain Portal 2 has the greatest narrative in all of video games - it’s expressed purely in gameplay, it’s funny as hell, and has a surprisingly deep emotional core.
It's not my favorite game, nor the favorite of anyone I know, but I'd be willing to guess it's worth top 3 (or at least 5) for almost everyone. Which is to say, in a first-past-the-post scenario it wouldn't win, but in a run-off-vote scenario it may very well be the best game ever made.
It's really not useful to make top n lists. As soon as there disagreement, distinctions and catagories evolve. Its probably sufficient to simple state you enjoyed the game, and to highlite some aspects you enjoyed.
This is, of course, a matter for debate, but as a huge Monkey Island fan, I think Portal 2 is funnier, at least in part because of the excellent delivery of the lines.
Regardless of which title you think is better, though, I think the morale is clear. Never pay more than $20 for a computer game.
So carefully selected video games are the real brain training. Who would have thought..
The following finding also continues to be interesting:
“More remarkably, we found that playing an action video game can virtually eliminate this gender difference in spatial attention and simultaneously decrease the gender disparity in mental rotation ability, a higher-level process in spatial cognition.”
https://www.ncbi.nlm.nih.gov/m/pubmed/17894600/
Some of this stuff is common sense—it’s amazing what length people go researching certain subjects. For instance, if your brain does a particular type of action (like tracking fast moving objects) it gets better at doing that in general. Same thing with studying math, people get better at solving problems.
Now, if those problems being solved in games or otherwise are limited in scope, well so are the gains achievable.
So, the question then is, how to APPLY that to a general set of problems. Not how it affects the brain doing a particular task. For instance, teach a man to fish salmon then ask him to take that knowledge and figure out how to catch whales or even plastic pellets in the ocean?
Or how about if one study math how can it increase ones art ability or vice-versa? It changes perceptions etc. Seems common sense to me I dunno.
> Some of this stuff is common sense—it’s amazing what length people go researching certain subjects.
Not sure how one can be science-minded and make statements like that. The entire point of scientific studies are to either verify the reasons for things that we may think are intuitive and also explore when that ends up falling apart (which is obviously when things get more interesting). Pretty sure it was "common sense" that the Earth was flat and that the Sun revolved around Earth not too long ago.
It's not obvious that it could eliminate the gender gap. If men start higher, why couldn't they stay in the lead as both sexes improved? That's why you have to do the research.
> So, the question then is, how to APPLY that to a general set of problems.
That's the tricky thing. As far as we know, IQ is the best proxy for general problem solving ability. From what I've read/heard from intelligence researcher; however, is that while you can train folks to excel at certain types of problems, it doesn't really improve one's general problem solving ability.
Think of an IQ test as a random sampling of problem solving exercises of various types from the ___domain of all possible problem solving exercises. Now imagine generating a whole bunch of such tests, and then using those tests on a bunch of people to setup your distribution (i.e. set the IQ scoring for each test). While it is true that in general the scores are going to be highly correlated across tests, each test is going to have a different composition in terms of the kinds of problems it presents. From what I understand, training may help on certain classes of problems (and thus on particular tests) but there isn't any training that will meaningfully improve the average of all your scores across all of the tests.
Anecdotal, but I completely agree that Portal supercharges your brain, and I've believed this for years. Back around 2011 when portal 2 came out we had a family reunion. I had been playing a lot of Portal at the time. One of my uncles broke out one of those brain teaser type metal ring puzzles where you have to pull the rings apart.
It made it's way around the room as we talked and no one could figure it out. As soon as it came to me, I saw clearly exactly what needed to happen to solve it, and within seconds it was solved. I still think it must have been all the Portal 2 I had been playing.
So to make real genius kids, throw away those Baby Einstein videos, ditch the "brain training" apps, and just play.
What would the play curriculum be?
Minecraft
Portal 1 & 2
The Talos Principle
Infinifactory
What else? I'm thinking something that involves a fourth spatial dimension or causality manipulation would also be good. I'd recommend Spellbreaker if I thought kids would tolerate text-only games these days.
The Incredible Machine, Lemmings, and just playing around with Logo were pretty awesome for problem solving fun, I think, back in my day. As for messing with causality, Braid comes to mind. To expand perspective, and for the sheer beauty of it, Gorogoa. Almost any RTS will give you a good feel for simple differential equations, combinatorics, economics, and a little game theory. I used to mess around with some incredible indie content on Kongregate, especially puzzle games, some of which were truly innovative and wonderful. (Too bad it is so hard to monetize awesome little indie games! Good luck to poki.com, itch.io, and their ilk!)
But the best foundation for the mind, I think, is to keep younger kids IRL, let them make up their own games, and keep it physical and outside as much as possible. Much of our human consciousness involves moving around abstract space, analogous to physical motion, and so learning to move and think IRL is vital to developing consciousness, wherever it may eventually play.
Maybe I'm just salty that I didn't figure out you could pick up laser redirectors without breaking the connection until after beating the game, but The Talos Principle always seems out of place in these lists. I've always seen it as a game focused on its story, which by necessity involves puzzle-based gameplay; but I've never thought that the puzzles were that good. In particular, it feels like the game actively discourages you from using 3D spatial reasoning, in favor of the puzzles being about using the tools you have to collect more tools. Most of the puzzles could have just been a 2D top-down perspective.
The puzzles needed for the low-effort ending are all like that, but some of the puzzles required for the 100%-completion ending are literally outside the box, requiring you to use tools from multiple discrete puzzle zones together, or to liberate tools from their zones to use in a portion of the level outside the puzzle fences.
Additionally, there are multiple Easter eggs peppering the game that take some amount of subversive reasoning to see or touch.
There's a "solve the puzzle as presented" level of reasoning, and there's a "break the puzzle" level of reasoning. Some puzzles have a "get the tetromino" solution, a "get the star" solution, and a "get somewhere I shouldn't be able to reach" solution.
For instance, have you found the Space Core, from Portal 2, in The Talos Principle?
If that's your list, I clearly need to check out The Talos Principle and Infinifactory. Minecraft and Portal are easily my son's favourite games. He seems to have lost interest in Roblox, Baba is You was fun but didn't really catch on.
He also enjoyed Braid, but many of the puzzles there were a bit too hard for him at the time. Maybe we should try again. And he's taken an interest in all the Supermario games now.
A neuroscientist friend of mine says Lumosity is not just junk science, but junk implementation of junk science. I would imagine just about any real puzzle solving task would have a greater impact on cognitive ability than the specific case of Lumosity.
I remember the ah-ha moment while interviewing at Lumosity was when the CEO burst in the room and excitedly told me how much he wanted me to make their site "Web 2.0". My heart sank and I noped out of there pretty quickly. A junk implementation of sound science sounds about right to me.
I think my only hope for learning Spanish at this point would be a game that starts out in a Texaslike setting, with the storyline missions sending the player into progressively more Spanish-speaking situations, with more Hispanic-specific cultural elements, until complete immersion occurs.
Start out ordering tacos for lunch. End by smuggling someone out of Venezuela via the Orinoco in a narco-submarine, with paramilitary groups ready to execute you if you mistranslate something.
Not surprised by this as Portals puzzles combine so many different abilities like abstract reasoning, planning, dexterity and reflexes. Now, the question is of course if other 3D video games (e.g. DOOM) have similar health benefits. I'd say Portal is a pretty "intellectual" game compared to most shooters, so I'd expect the overall effect to vary. Navigation in three-dimensional environments seems to be quite a complex task in itself though, so maybe that's already where most of the effect is coming from?
Oh, there are more. Out of my mind, I can cite The Talos Principle, Qube and Antichamber.
Maybe in slightly different categories: Fez (~2D), Mirror's edge (parkour), infinifactory. The Stanley Parable as well, perhaps?
Then there are numerous 2D ones, and I'm pretty sure I am forgetting some that I have in my library. Portal isn't unique, but it has one of the most dynamic and entertaining narratives. It also spearheaded a whole lot of "portal clones", which was nice to see, as I quite like those puzzle games.
From your first three I've only seen Antichamber. I agree that it's a cognitive challenge like portal, but I also think it doesn't involve the same kind of spatial reasoning portal does. Well, at least not in the (otherwise ubiquitous) 3D space.
Well, the most difficult part of Antichamber is that the 3D space is a non-euclidian one, so you have to adjust the way you think space. But I still think that the main game element is how to find your way from point A to point B in that space. And pathfinding isn't a trivial exercise in euclidian space, already, which is the point of the article.
I recommend the Talos Principle, even though I haven't finished it yet. Some challenges are quite hard, and it has even more spatial reasoning than portal. Here also, enigmas revolve around getting you from point A to point B, but you usually have to carry along some tools to help you. This often ends up being a variant of the Hanoi tower, or the wolf, goat and cabbage, but in a space with more than one dimension, unlike those classic riddles. And there is some lore hidden in the game world that can be quite interesting to puzzle together.
Now, if you want more portal, the game modding community is quite profusive around Valve's IP. You have Portal stories: Mel, Aperture tag, Portal: prelude. The quality is often worse than complete games, but the puzzles can be quite challenging.
I haven't played Portal, but, based on what I've seen of Portal, Human Fall Flat may be similar. You solve puzzles in a 3d world to get through stages of the game. It's pretty fun.
The main advantage I'd expect Portal to have over other games is the spatial memory required to remember where you've already put your portals, while they're not currently on the screen. I expect that a linear game would score the lowest, since there's no incentive to build a mental model of the map inside your head. Games with sprawling levels that encourage backtracking would be somewhere in between, but those are kind of frustrating and outdated these days, and you can still complete them by randomly walking around until you reach your next goal. Portal games have backtracking, but you need to intentionally place a portal in both locations, which you can't do accidentally.
I don't know any of those but could it be that Lumosity is just harder on your brain making you more tired and mentally exhausted? I mean, if a group relaxes for 8 hours and other group does lumosity for 8 hrs, why should the second group do better after exhausting their brains for so long?
Good question, I think there was a break between the training and the posttest, but it isn't quite clear from the description in the paper:
"Each participant spent 10 h in the study, which spanned four separate sessions in the on-campus laboratory of the university, across 1-2 weeks. Each of the first three sessions lasted 3 h. Session 4 lasted one hour - solely for administering the posttest battery. At the beginning of the first session, subjects in both conditions were asked to read and sign the consent form. After signing the form, subjects were randomly assigned to one of the two conditions. All subjects were then administered an online set of pretests. Once they finished the pretests, they logged in to play their assigned game for the rest of the first session. They continued to play their assigned game for the entire second and third session (for a total of 8 h of gameplay). During the last session, they completed the online set of posttests."
"Before and after 8 h of gameplay, subjects completed an online pretest and posttest battery of tests measuring our focal constructs and their relevant facets."
Sounds like the post-test was done separately to me.
Session 1 is probably about 2 hours of "training", with the rest of the time eaten up by the pretest and consent paperwork. Sessions 2 and 3 would be be almost entirely training (3 hrs each), which gets you to 8 hours of training in 3 sessions, leaving the last one for just the post-test.
Could it be that Portal players felt they were more successful in their games than Lumosity players felt?
I know that successfully completing a task(s) just before a test will result in better performance on tests, and failing will result in poorer performance.
They aren't asking how a player felt about Portal or Luminosity. It looks like they did some standardized tests after they played the respective games.
Yes, but parent is arguing that the feeling could be what is affecting the test results more than any cognitive abilities or enhancements as a result of playing the game. I.e. if you "won" the Lumosity games (which normally get harder until you fail) and "lost" at Portal, is the effect the same or is it reversed?
They did ask how the players felt, because the GP has a good point, and included it in the analysis:
"3.4.3. Enjoyment item
After participants in each condition finished their 8 h of gameplay and before the posttest battery of tests was administered, we presented an item to players concerning their enjoyment of the assigned game. This was a single self-reported response to the statement: “I enjoyed playing (Portal 2 or Lumosity).” The 5-point Likert scale ranged from 1 (strongly disagree) to 5 (strongly agree)."
"The measures were controlled for player enjoyment in all models (when significant) to rule out any effect enjoyment might have in learning gains. That is, based on participants' responses to the
statement about enjoying their assigned game, for Portal 2 participants, enjoyment M = 4.32; SD = 0.93, while for the Lumosity participants,
M = 3.50; SD = 1.05. The difference between the two groups' enjoyment is significant, with a strong advantage for the Portal 2 group: F(1,73) = 12.69; p < .001. Cohen's d = .83, which is a large effect size."
"Enjoyment was a significant covariate in the VSNA ANCOVA model"
I forget the behavior / explanation but if you take two groups and have one successfully complete tasks, and another unsuccessfully, and then test them both you'll find the folks who just previous to the test successfully completed tasks will do better than those who were unsuccessful.
This isn't true. 8 hours says something about an 8 hour activity or event. Which can hint at a possible connection to an annual activity or event.
Scientists do not have unlimited resources or sample sizes that are unlimited. They make do with what they can and an 8 hour study does produce results worth looking at. No logic says otherwise.
Most studies probably use much less than 8 hours of a respondent's time, I believe. The greater issue is the use of only 77 test subjects to begin with.
I wonder if any of this boost, or even a significant portion, comes from just being happier and relaxing after playing a game for 8 hours. Rather than some perceived brain training.
People who have played a funny, pretty looking and relaxing puzzle game prior to solving skill based questions perform well. What have we learned? Was it the puzzles or GlaDOS?
Maybe all GlaDOS' subtle (and not so subtle) digs at humanity encourages humans to do their best.
Or maybe the humorous absurdity of a testing-obsessed "abusive mom" computer makes people relax more in actual testing situations. I suspect these sorts of aptitude tests get a lot of their power from the intimidation factor.
I would self-immolate for a game that simultaneously served as the sequel to both HL2 and Portal 2 (i.e. Gordon Freeman has both a gravity gun and a portal gun)
Wow, they totally fell under my radar, which is impressive when they own the #1 game distribution platform in the world. I remember seeing mention of Artifact, but didn't notice DOTA at all.
Valve have stated over and over that teleport locomotion is what they believe is the correct one. Portal's mechanic is about movement as much as portals - it simply wouldn't work in this environment.
The Portal/Half-Life universe (they're the same) itself? Absolutely.
They have set such an impressively high bar it would likely be a disappointment. I am a fan of letting things end on a high note instead of turning them into a never ending franchise and trying to extract every last cent out of the IP.
> In this study, we tested 77 undergraduates who were randomly assigned to play either a popular video game (Portal 2) or a popular brain training game (Lumosity) for 8 h.
Sorry my comment should have been clearer. I wasn't actually asking for the sample size - i was making a statement that the sample size was ridiculously small as to be irrelevant. It was more like an eye roll. Thank you for responding to my literal comment snd reminding me to be clearer in my messages :)
77 is a large enough sample size to produce statistically significant results in a lot of cases, depending on the problem. For example, using 77 men and women, you could easily see a statistically significant difference in height between genders.
This paper is pretty simple, it's just a straight-forward before-and-after test. They are holding a lot of factors constant and not doing an observational study, so 77 is actually a fine sample size for something like this.
There's a power analysis in the paper showing that they've got plenty of statistical power. (Estimated N=34 subjects for 90% power, but they used 77 subjects instead).
